Blog

Rare Consulting recently compared the life cycle greenhouse gas emissions associated with different forms of fuels for passenger vehicles. A number of fuel types and their emissions were compared over the full life cycle, which includes the extraction of raw product (such as oil or gas), refining, transportation to consumer and then combustion emissions in a passenger vehicle.

The analysis raised questions regarding the assumptions that lie behind some modelling, and showed that greenhouse gas emissions assessment is a difficult and uncertain process. In particular, the results appeared to conflict with those for electric vehicles and greenhouse gas emissions presented in the Garnaut climate change review. The review concluded that based on the average emissions intensity of the Australian grids, an electric car today would generate about 30% more emissions than a similar petrol-fuelled car, a figure that was at odds with the results of Rare’s analysis (although in an earlier blog entry we did refer to this analysis in good faith).

Given that there are no figures or assumptions behind this conclusion, Rare looked more closely at the Garnaut numbers. Using combustion figures from the NGA factors workbook, and an average passenger car fuel economy of 11.5 L per 100 km from theABS 2007 Survey of motor vehicles it was possible to conclude that the average petrol-fuelled passenger vehicle in Australia will emit approximately 270 g CO₂-e per kilometre travelled at the tailpipe. Note that this figure does not include those emissions arising from the exploration and extraction of crude oil, distribution to refineries and refining, and shipping to the point of consumption. Therefore it does not encompass the full life cycle of emissions. However, for the purpose of this exercise only the combustion figure was used.

Looking at the emissions produced by a typical mid-sized passenger electric vehicle travelling one kilometre, and fuelled by the Australian grid, a number of key assumptions were made: the average Australian electricity generation emissions intensity is approximately 860 g CO₂-e/kWh (IEA 2007), and a vehicle of this size requires 0.2 kWh from the battery to the wheels per kilometre travelled. Although this figure is considerably less for purpose-built small electric vehicles, there are also many conventional small passenger cars that do not consume 11.5 L per 100 km, and we were seeking to analyse a comparable vehicle in this calculation.

There were also a number of efficiency losses to be considered between the generation of electricity and the wheels of the vehicle: grid transmission losses (10%), charging losses (5%), and battery storage losses (5%), based on a trickle charging application.

Applying these figures to the energy requirements of the vehicle, it was estimated that per kilometre travelled, a mid-sized electric vehicle requires 0.246 kWh of electricity to be generated. At the average emissions intensity of 860 g CO₂-e/kWh, this equates to 212 g CO₂-e per kilometre travelled, or about 20% less than an average Australian petrol-fuelled passenger car.

Note that this analysis was highly conservative, and did not consider the upstream emissions associated with producing and delivering the liquid fuel to a petrol-fuelled car. If these emissions had been considered, and if like boundaries had been adopted for the life cycle analysis, it is reasonable to assume that an electric vehicle would appear even less emissions intensive on a relative basis.

To provide a more comprehensive assessment, we also considered a fast charging application in which there are greater efficiency losses to be considered at the point of charging (10%) and battery storage (20%). As the charging current goes up, the internal heat generated in the battery goes up and the internal resistance climbs, meaning a higher proportion of the energy gets lost as heat. Under this scenario the electric vehicle requires 0.308 kWh of electricity to be generated to travel one kilometre, producing 264 g CO₂-e per kilometre travelled, still less than an average Australian conventional passenger car.

So where do the Garnaut Review numbers come from? Without being privy to the assumptions behind the conclusions, we can at best assume that the Review selected small cars with very low fuel consumption and hence lower emissions intensity to conduct its unfavourable comparison with electric vehicles.

However, life cycle analysis is complex and open to significant levels of ambiguity and even tailoring to one’s desired outcomes. At the very least, it demonstrates that there is a need to question the assumptions and numbers behind emissions modelling, even at the highest level.